The research project DACSEIS provides an European quality-management-system for socio-economic data in the national statistical sectors obtained by complex survey sampling. A team of international experts, working on applied and theoretical statistics, was formed to analyse and solve these problems.
Heart of the problem of quality measurement of statistical data are variance estimation methods in complex multi-purpose sampling. Complex samples are currently used in all national official and non-official statistic institutes to guarantee a broad socio-economic database. For the future of the new economy in Europe it is going to be important to provide a harmonised database that can be evaluated according to consistent quality standards.
Based on these data, new indicators for the rating of aggregate economic development in Europe are formed. The major aim of the new standards for precise statistical information is the consistent classification, methodological unification, basic evaluation, and improvement of existing methods for variance estimation in complex survey samples.
The methods, that are to be investigated, include resembling methods, methods for unequal probability designs, combining register and survey data, raking adjustment methods, and variance estimation methods for change. Further estimation procedures that face on special topics, e.g. the Small Area Estimation, are examined.
Finally, a catalogue of user-oriented instructions and criteria for all statistical institutes and other professional users of these methods will be provided. The computer codes for the relevant variance estimation methods investigated in the project will enable the statisticians to use new tools for quality control. The distribution of such an innovative and reliable methodology for quality management of statistical data is a central aim of the project. Additionally, standard software packages for survey sampling are examined and evaluated.
The evaluation includes the implementation of standard and non-standard variance estimation methods for complex surveys. The possibility of an enhancement of further variance estimation methods into the relevant software packages will also be checked.
All relevant methods are to be inspected, analysed, and evaluated with respect to their applicability to complex surveys. This is going to be achieved by simulated but realistic complex universes as a basis for an detailed Monte-Carlo simulation study of the surveying methods and its variance estimation.
In particular, their specific efficiency in extreme or extraordinary situations is to be tested using a computational approach. The project will finally provide an extensive documentation containing the catalogue of best practice recommendations in order to guarantee that users are able to apply the adequate estimation methods for complex surveys, either official or non-official applications.
The exchange of information with end-users will be achieved by discussion groups in strong connection with Eurostat.
The dissemination of the DACSEIS research will be granted by the different traditional publication procedures, a final international conference, and a public WWW-server, that is installed at the beginning of the project.